Skip to content

talk

Overview

An R-package for analyzing natural language with transformers-based large language models. The talk package is part of the R Language Analysis Suite, including talk, text and topics.

  • talk transforms voice recordings into text, audio features, or embeddings.

  • text provide many language tasks such as converting digital text into word embeddings.

    talk and text provide access to Large Language Models from Hugging Face.

  • topics visualizes language patterns into topics to generate psychological insights.



The R Language Analysis Suite is created through a collaboration between psychology and computer science to address research needs and ensure state-of-the-art techniques. The suite is continuously tested on Ubuntu, Mac OS and Windows using the latest stable R version.

📣 Online Workshop: Analysing Human Language using R

August 11–13, 2026 — three half-day sessions (2:00–5:00 pm CEST / 8:00–11:00 am ET)

Learn the full R Language Analysis Suite — talk, text, and topics — with Oscar Kjell.

Read more & register →

Point solution for transforming talk to embeddings

Recent advances in speech and language modelling have dramatically improved how well computers can transcribe and represent human speech. The talk-package makes these state-of-the-art models — such as Whisper from HuggingFace — easily accessible from R. All analyses run locally on your own computer, so recordings never leave your machine — important when working with sensitive data such as clinical or research interviews.

With talk you can:

  • Transcribe speech to text: turn audio recordings into text, in many languages.
  • Identify who says what and when: split a conversation by speaker, with time-stamped speaker turns (diarisation).
  • Transform speech into embeddings: represent whole recordings — or each speaker turn — as numeric vectors that capture acoustic and psychologically relevant properties of the voice, beyond the words themselves.
  • Use the embeddings in downstream analyses: for example, training models to predict psychological constructs with the text-package.

See the Getting started tutorial for a walk-through of the workflow, and HuggingFace for the available models.

GitHub